Antitrypanosomal activity of 5-nitro-2-aminothiazole-based compounds

A small series of 5-nitro-2-aminothiazole-based amides containing arylpiperazine-, biphenyl- or aryloxyphenyl groups in their core were synthesized and evaluated as antitrypanosomatid agents. All tested compounds were active or moderately active against Trypanosoma cruzi amastigotes in infected L6 cells and Trypanosoma brucei brucei, four of eleven compounds were moderately active against Leishmania donovani axenic parasites while none were deemed active against T. brucei rhodesiense. For the most active/moderately active compounds a moderate selectivity against each parasite was observed. There was good correlation between lipophilicity (clogP value) and antileishmanial activity or toxicity against L6 cells. Similarly, good correlation existed between clogP values and IC50 values against T. cruzi in structurally related subgroups of compounds. Three compounds were more potent as antichagasic agents than benznidazole but were not activated by the type I nitrorectusase (NTR).


Introduction
American trypanosomiasis (Chagas disease), human African trypanosomiasis (HAT or sleeping sickness) and leishmaniasis are considered neglected tropical diseases (NTD) and represent a severe global health problem [1,2]. It is estimated that together these three diseases, caused by protozoan parasitic infections, affect approximately 20 million people and are responsible for more than 110,000 deaths annually [2]. African trypanosomiasis is endemic in many sub-Saharan African countries and is caused by Trypanosoma brucei rhodesiense and T. brucei gambiense. Chagas disease, caused by Trypanosoma cruzi, is endemic in South and Central America but is now spreading worldwide, mainly due to human and vector migration [3,4]. Leishmaniasis, caused by more than 20 Leishmania species, occurs throughout tropical and sub-tropical regions and is now spreading worldwide as an HIV co-infection [5].
Treatment of these NTD is currently based on a series of problematic drugs. Thus, nifurtimox (Nfx) and benznidazole (Bnz), the two currently used medications for Chagas disease are associated with limited efficacy, severe toxicity and long treatment requirements [6,7]. Similarly, drugs used to treat HAT and leishmaniasis are highly toxic (e.g. melarsoprol, antimonials), may require i.v. administration (e.g. melarsoprol, suramin, DFMO, antimonials), can cause severe side effects, or are of high cost (e.g. DFMO, liposomal amphotericin B, miltefosine and paromomycin) [8e10]. Therefore, there is an urgent need for new effective, safe and affordable alternatives.
We have shown that several chemical classes of 3-nitro-1H-1,2,4-triazole-based compounds exhibit excellent antichagasic activity both in vitro and in vivo [18e25]. Furthermore, appreciable anti-HAT activity was also observed in vitro with several such analogs [18e25] whereas in vitro antileishmanial activity was demonstrated with a sub-class of 3-nitrotriazole-and 2nitroimidazole-based aryloxyphenylamides [25]. Nitro-activation by an oxygen-insensitive type I nitroreductase (NTR), an enzyme located in the mitochondrion of trypanosomatids and absent from most other eukaryotes, is partially responsible for the trypanocidal activity of these and other nitroheterocyclic compounds [18,19,21e29]. More recently, we have synthesized 3-nitrotriazolebased rigid amides and carbinols which act as bifunctional agents; they exert their antitrypanosomal activity upon activation by type I NTRs and by inhibiting the parasite's CYP51 enzyme [23,25]. Interestingly, 3-nitrotriazole-based compounds are significantly more potent and less toxic than their 2-nitroimidazole-based counterparts [18e25,30].
Here we have expanded our research by investigating the role that another nitroheterocyclic ring, 5-nitro-2-aminothiazole, plays in antitrypanosomatid activity. Nitrothiazole-and nitrobenzothiazole-containing compounds exhibit antiparasitic, antibacterial, antifungal and antitubercular activities [31e34]. Therefore, we have synthesized and evaluated in vitro a small series of 5-nitro-2-aminothiazole-based compounds bearing moieties that were previously proven effective in the trypanocidal activity of 3-nitrotriazole-based agents.

Chemistry
The synthesis of 5-nitro-2-aminothiazole-based compounds ( Table 1) is straightforward and based on well-established chemistry, outlined in Scheme 1.
The precursor alkylchloride 1 as well as compound 8 were formed by nucleophilic substitution of 2-chloroacetyl chloride and [1,1 0 -biphenyl]-4-carbonyl chloride, respectively, by 5-nitro-2aminothiazole, in the presence of triethylamine. Amides 2e7 were obtained by nucleophilic substitution of alkylchloride 1 by an appropriate piperazine at room temperature and in the presence of triethylamine. Finally, amides 9e12 were prepared by nucleophilic substitution of alkylchloride 1 by the potassium salt of an appropriate phenol in DMF, by heating for 3e4 h at 60 C. Efforts were made to improve the yield of amides 9e12 by changing the solvent to anhydrous DMSO or CH 3 CN without any positive results. All final compounds and intermediates were characterized by 1 H NMR (500 or 400 MHz) and HRMS.

Antiparasitic activity and cytotoxicity
Compounds in Table 1 were screened for antiparasitic activity against three trypanosomatids: T. cruzi, T. b. rhodesiense and Leishmania donovani. The concentration of compound that inhibits parasite growth by 50% (IC 50 ) was calculated from dose response curves for each parasite (Table 1). In addition, compounds were tested for toxicity in L6 rat skeletal myoblasts, used as host cells for T. cruzi amastigotes, in order to calculate a selectivity index for each parasite (SI ¼ IC 50L6 /IC 50parasite ) ( Table 1). Antiparasitic activity was evaluated according to the following criteria: an IC 50 of <4.0 mM, between 4.0 and 60 mM or >60 mM, designates 'active', 'moderately active' or 'inactive' compounds, respectively, against T. cruzi amastigotes; for blood stream form (BSF) T. b. rhodesiense, IC 50 values of <0.5 mM, between 0.5 and 6.0 mM or >6.0 mM identify 'active', 'moderately active' or 'inactive' compounds, respectively; finally, for L. donovani amastigotes, IC 50 of <1 mM, between 1.0 and 6.0 mM or >6.0 mM, provides 'active', 'moderately active' or 'inactive' compounds, respectively [35].
According to the criteria set above, all tested compounds in Table 1 were active or moderately active antichagasic agents (green or light green, respectively). Four compounds (6, 9, 10 and 12) were moderately active (light green) against L. donovani parasites whereas no compound demonstrated antiparasitic activity against T. b. rhodesiense. Moreover, all compounds showed PSA values > 100 Å 2 , which makes them highly unlikely to be capable of penetrating the bloodebrain barrier and demonstrate anti-HAT activity in vivo.
Several analogs (3, 5e9) demonstrated IC 50 values < 50 mM against L6 host cells, presumably due to their high lipophilicity (Table 1), resulting in low selectivity indices. However, even compounds with IC 50 values > 50 mM against L6 cells demonstrated a less than ideal SI, which is desired to be ! 50 for T. cruzi and !20 for L. donovani [35].
Another general observation is that the 5-nitro-2aminothiazole-based amides are significantly more lipophilic (Table 1) than their 3-nitrotriazole-based counterparts with the latter having clogP values between À0.198 and 3.1. In addition, 5nitro-2-aminothiazole-based amides demonstrate higher PSA values than their 3-nitrotriazole-based analogs (the latter having PSA values less than 116 [24]), which may negatively affect cell permeation [36]. These features may contribute to the higher toxicity of the nitroaminothiazoles in L6 cells and their reduced potency against the parasites (Table 1).
There was an excellent correlation between antichagasic activity and lipophilicity (R 2 ¼ 0.979) in the piperazine-amide subgroup of compounds 3e7 (which were active against T. cruzi) as shown in Fig. 1. Thus, the piperazine amide 6 with the highest clogP value (4.51) was the most active compound against T. cruzi, demonstrating an IC 50 of 571 nM, 3.86-fold more active than Bnz (Table 1). Compound 6, however, was about 8-fold less active than its 3nitrotriazole-based analog, in which the piperazinic ring is directly connected with the carbonyl (piperazide) and the nitrotriazole ring is connected with the carbonyl through a methylene group [24]. SAR follows the same rules observed in the 3nitrotriazole-based piperazines and piperazides [20,24]. Therefore, dichlorophenylpiperazine 5 was a slightly better antichagasic agent than trifluoromethylphenylpiperazine 3, the latter demonstrating better antichagasic activity than the methoxyphenylpiperazine 4 or the heteroarylpiperazine 7. It is not clear if the electronic effect of substitution plays a role in activity other than influencing the clogP value.
With regard to in the structurally related compounds 8e12, there was also excellent correlation between antichagasic activity and lipophilicity (R 2 ¼ 0.977), with the most lipophilic biphenylamide 8 having the lowest IC 50 value against T. cruzi (Fig. 2). Interestingly, and despite their relatively high lipophilicity, the (phenoxy/phenyl)phenoxy-derivatives 9e12 were only moderately Table 1 In vitro antiparasitic activity, host toxicity and physical properties of 5-nitro-2-amino-thiazole-based amides.

Analysis of anti-Leishmania activity
There was no correlation between antichagasic and antileishmanial activity for compounds 2e12. Thus, compounds 9, 10 and 12 that displayed a moderate antichagasic activity (IC 50 values of 23e35 mM) demonstrated high potency towards L. donovani, yielding IC 50 s of 4.27e5.40 mM.
In contrast to antichagasic activity and, with the exception of the most lipophilic compound 6, the 3-nitro-2-aminothiazole-based piperazine amides 2e7 were generally less active antileishmanial agents than the (phenoxy/phenyl)phenoxy derivatives 9e12. With regard to the role of substitution in the antileishmanial activity, the same rules existed, which were mentioned above for antichagasic activity. There was good correlation between clogP and logIC 50 values against L. donovani parasites for all compounds ( Table 1), regardless of activity (Fig. 3). Therefore, once again, lipophilicity was very important determinant for antileishmanial activity. However, lipophilicity also resulted in relatively high toxicity with good correlation between clogP values and logIC 50 values in L6 cells (Fig. 4).  Table 1.

The role of type I nitroreductases
To elucidate the mechanism of action of the novel compounds in Table 1, representative derivatives (3, 5, 6, 7, 8 and 11) were evaluated as substrates of purified, recombinant TbNTR and their enzyme specific activity, measured as nmol NADH oxidized min À1 mg À1 protein, compared with that of benznidazole (Fig. 5). With the exception of compound 3, all other tested analogs were deemed to be poor TbNTR substrates, consistent with results obtained for previously studied N-substituted 5-nitro-2aminothiazoles where the substituent had electron withdrawing groups [37]. Compound 3, which was not the most potent analog against T. cruzi amastigotes, provided a similar TbNTR specific activity to BNZ (Fig. 5).
To further determine whether NTR plays a role in metabolizing the substrates within the parasite, the above subset of compounds were phenotypically screened against BSF T. b. brucei expressing wild type or elevated levels of TbNTR. Compound 7, with an IC 50 > 10 mM in wild type parasites was not screened against the recombinant line. For most of the remaining compounds cells overexpressing TbNTR were only 2-to 3-fold more sensitivity than controls to the agent, a relatively low shift when compared to that observed with nifurtimox. The biochemical and phenotypic screening data suggest that NTR plays little or no role in the metabolism of these compounds within the parasite itself. As these 'non-TbNTR' activated compounds display moderate growth inhibitory properties towards wild type T. b. brucei ( Table 2) the mode of action of these compounds remains unknown although this antitrypanosomal activity does appear to be sub-species specific: The selected 5-nitrothiazole-based compounds tested are up to 12-fold more potent towards T. b. brucei than against T. b. rhodesiense.
In conclusion, novel N-substituted 5-nitro-2-aminothiazoles with an arylpiperazine-, biphenyl-or aryloxyphenyl group in the core were active or moderately active antichagasic agents and moderately active against T. b. brucei parasites. Only one derivative, compound 6, demonstrated activity against T. cruzi amastigotes at nM concentrations and was about 4-fold more potent than BNZ. In addition, some of these compounds demonstrated a moderate antileishmanial activity against L. donovani axenic amastigotes. These particular compounds were not good substrates for type I NTR.
However, more simple N-substituted 5-nitro-2aminothiazoles were shown to be excellent substrates of type I NTR and their antiparasitic activity was increased about 10-fold in NTR overexpressing T. b. brucei [37]. Interestingly, these more simple N-acyl substituted 5-nitro-2-aminothiazoles demonstrated significantly lower clogP values than the compounds described in here [37]. Therefore, this class of compounds deserves further investigation and structural optimization may provide leads for further development.

General
All starting materials and solvents were purchased from Sig-maeAldrich (Milwaukee, WI), were of research-grade quality and used without further purification. Solvents used were anhydrous and the reactions were carried out under a nitrogen atmosphere and exclusion of moisture. Melting points were determined by using a Mel-Temp II Laboratory Devices apparatus (Holliston, MA) and are uncorrected. Proton NMR spectra were obtained on a Varian Inova-500 or an Agilent Hg-400 spectrometer at 500 or 400 MHz, respectively, and are referenced to Me 4 Si or to the corresponding solvent, if the solvent was not CDCl 3 . High-resolution electrospray ionization (HRESIMS) mass spectra were obtained on a Agilent 6210 LC-TOF mass spectrometer at 11,000 resolution. Thin-layer chromatography was carried out on aluminum oxide N/ UV 254 or polygram silica gel G/UV 254 coated plates (0.2 mm, Analtech, Newark, DE). Chromatography was carried out on preparative TLC alumina GF (1000 microns) or silica gel GF (1500 microns) plates (Analtech). All final compounds were purified by preparative TLC chromatography on silica gel or alumina plates and also checked by HPLC (!95% purity).  Table 1. Table 1.

General synthesis of N-(5-nitrothiazol-2-yl)acetamides 2e7
A dichloromethane solution (6 mL) of an appropriate piperazine (1 eq) and triethylamine (3 eq) was added drop wise to a suspension of 2-chloro-N-(5-nitrothiazol-2-yl)acetamide (1) in 5 mL CH 2 Cl 2 and the reaction was kept at room temperature under a nitrogen atmosphere and stirring for 48 h. The reaction solvent was evaporated and the residue was redissolved in ethyl acetate. The inorganic salts were filtered away and the residue was chromatographed on preparative TLC silica gel plates with ethyl acetate/petroleum ether as eluent to obtain the desired pure product as a powder or crystals. Purity was also checked by HPLC and it was !95%.

General synthesis of N-(5-nitrothiazol-2-yl)acetamides 9e12
An appropriate phenol (1.05 eq) and K 2 CO 3 (2.1 eq) were mixed together in dry DMF (4 mL) and stirred for an hour. Then a DMF solution (4 mL) of 2-chloro-N-(5-nitrothiazol-2-yl)acetamide (1 eq) was added through a funnel and the reaction mixture was heated at 60 C for 3e4 h. The solvent was evaporated and the residue was chromatographed on silica gel preparative TLC plates using ethyl acetate: petroleum ether mixture as an eluent. The desired product is formed in relatively small yield ( 35%) and appears immediately after the unreacted phenol on TLC. Changing the solvent to CH 3 CN or DMSO did not improve the yield.   In vitro activity against T. cruzi, T. b. rhodesiense, L. donovani and cytotoxicity assessment using L6 cells (rat skeletal myoblasts) was determined using a 96-well plate format as previously described [38]. Data were analyzed with the graphic program Softmax Pro (Molecular Devices, Sunnyvale, CA, USA), which calculated IC 50 values by linear regression from the sigmoidal dose inhibition curves.

In vitro T. brucei brucei antiproliferating assays and susceptibility studies
T. brucei brucei bloodstream form parasites were seeded at 1 Â 10 3 mL À1 in 200 mL of growth medium containing different concentrations of a nitrothiazole or nifurtimox. Where appropriate, induction of the TbNTR was carried out by adding tetracycline (1 mg/mL). After incubation for 3 days at 37 C, resazurin (2.5 mg per well) was added to each well and the plates incubated for a further 8 h. The cell density of each culture was determined as described before [26] and the IC 50 established.

Enzymatic activity studies with type I NTRs
Recombinant TbNTR was prepared and assayed as previously described [39,40]. The activity of purified his-tagged TbNTR was assessed spectrophotometrically at 340 nm using various nitrothiazole substrates (100 mM) and NADH (100 mM) with the enzyme specific activity expressed as nmol NADH oxidized min À1 mg À1 of enzyme. Benznidazole was used as control substrate.